D. Kindell1, S. Markowiak1, F. C. Brunicardi1, M. Nazzal1 1The University of Toledo College of Medicine,Department Of Surgery,Toledo, OH, USA
Introduction: In FY2014 Medicare began modifying payments to hospitals based on hospital heart failure (HF), acute myocardial infarction (AMI), and pneumonia (PNA) 30-day mortality rates through the Value-Based Purchasing (VBP) Program. To date, approximately 1.86 billion dollars in Medicare funding has been allocated to hospitals based on these scores. The purpose of this study was to determine whether these mortality rates were influenced by social determinants of health (SDOH).
Methods: Data were gathered from Centers for Medicare and Medicaid Services (CMS) and US Census Bureau archives. We created a database pairing individual hospital mortality outcome performance scores (30-day HF, AMI, PNA) to corresponding census measures at the county level. Pearson’s Correlation Coefficient (Pearson r) was used to test 133 SDOH against mortality outcomes. Student’s t-test was used to compare the top and bottom quartile of performers across each mortality outcome. The United States Department of Agriculture criteria were used to classify urban vs rural counties.
Results: In total, 2955 hospitals across 977 counties were analyzed. Greater rural setting for hospitals was correlated with worse 30-day mortality rates in heart failure (r= – .144), pneumonia (r= – .136), AMI (r= -.082, p <0.001). The top correlations associated with improved mortality rates were improved PNA mortality rates in communities with more college graduates (r=0.170, p<0001), better heart failure rates in communities with more immigrants (r=0.206, p<0.001), and better pneumonia performance rates in communities with more white-collar jobs (r=0.170, p<0.001). Predominantly Caucasian communities were correlated with worse HF outcome (r=-0.176, p<0.001), whereas lower income was correlated with worse PNA performance (r=-0.160, p<0.001). Although there were many significant correlations (p<0.001) between community characteristics and hospital AMI mortality, none were strong.
Conclusion: Of the SDOH analyzed, 79% demonstrated statistical significance within the AMI mortality outcome, 88% in the heart failure mortality outcome and 87.3% in the pneumonia mortality outcome. The weaker strength of correlations between SDOH and AMI mortality rates may be due to the decades old efforts by the American Heart Association to increase rural and poor hospital access to fast percutaneous coronary revascularization. Due to the disparities in SDOH between urban and rural hospitals, mortality outcomes will shift CMS incentive payments away from hospitals in disadvantaged communities.